AUDIO over IP - REDNET 3 & 16 Review. AES67 Sets A New Standard for Computer Audio

[jabbr]

Clear, AOIP must be considered as a audio driver. The issue is the lack of drivers on Linux platforms. I saw only drivers for Windows or Apple OS. Perhaps it will be soon...

Well, nearly..

AOIP is about the transport from your playback device to the DAC or incase of the Rednets, a digital interface just before the DAC.

 

[jelt2359]

 
 
Just received the Cybershaft Premium OCXO 10Mhz clock today. Product has nice solid build and comes with certificate of calibration with individual graphs (Allan Deviation and Phase Noise). The recommendation is  (like the Yggy)  to leave the clock on at all times if possible. I will let the clock run for a few hours before doing any listening.  No, I don't normally stack all my equipment like this but this was the best way to get everything in one shot. I am connecting the 10Mhz Cybershaft into the Liveclock via .5M 50-ohm BNC Cable and then running 2 1M 75-ohm cables from Liveclock to RedNet and Mutec. The Atomic clock indicator is lit up on the front  Liveclock display as well as in the Liveclock software which is a good sign that things seem to be working properly. Will provide additional feedback later after some listening sessions.

Can the 10Mhz also directly be connected to Mutec? I wonder how that will work, with Mutec then also be a wclk output to the Rednet.

 

[jabbr]

Can the 10Mhz also directly be connected to Mutec? I wonder how that will work, with Mutec then also be a wclk output to the Rednet.

Yes, the Mutec will accept a 10MHz reference clock to support its internal word clock generator.
Would indeed be interesting to see how the internal word clock + 10 MHz compares to the Antelope as external word clock for the Mutec.

Everthing depends on the design of each of the PLL and oscilator circuits.
The Grimm white papers are an interesting read, to give you an understanding when external clocks might or might not be helpfull to the clock stability.

Cheers

 

[gldgate]

As many here have done, I purchased an Antelope Liveclock and found it an improvement over internal RedNet and Mutec clocks. I was an external clock newbie so I purchased the Liveclock without any expectations. Given the positive outcome, I became interested in what a 10M clock could do. I know there is no concensus in regards to 10M. On the one hand you have testimony from some studio producers saying they can't live without it. On the other hand you have folks saying it makes no difference what so ever and it might as well be an expensive paper weight. Though I had interest, the expense was prohibitive. The Antelope 10M is over $5K, the Stanford Perf 10 is $3500 and the upcoming Mutec Ref 10 has been rumored to be 3000 Euros. All over my comfort zone financially. A little additional searching brought forth a potential alternative - Cybershaft of Japan. A few US based posters have had positive results (mostly feeding esoteric players) but the sample size is small with majority of users in Japan.
 
I initially contacted Hasegawa-san and inquired about the various 10M options (He sells both OCXO and Rubidium and even a unit that contains both). He asked what I was going to use the clock with and my musical preferences. He first confirmed that his clocks are compatible with the Antelope Liveclock and then mentioned that his recommendation would be the OCXO Premium for my listening habits (weighted more toward Classical side of things). Payment for clock and 50-ohm bnc cable was via Paypal and unit was shipped out via EMS in a few days. Total charge for clock, cable and shipping with exchange rate was $1K.
 
Unit arrived yesterday. The OCXO Premium has some real heft to it. It's easily 2x-3x times the weight of the Liveclock. The Cybershaft also has a better overall build quality than the Antelope. As mentioned in prior posts, the unit comes with certificate of calibration that shows individual unit test results with some graphs.
 
The Cybershaft was connected to the Antelope and the Atomic Clock connection of the Liveclock instantly lit up. The Atomic indicator also was indicated in the Antelope Software. I let the unit warm up for for the afternoon and early evening and started listening in earnest around 8PM. I wrapped around 2AM.
 
Overall impression is that Cybershaft into Liveclock is preferable to Liveclock on its own. Imaging is more solid and individual musical lines in a complex orchestral piece seem easier to distinguish/follow. One of my  "go to" evaluation recordings is Debussy's three nocturnes conducted by Haitink and the Concertgebouw on Philips. The nocturnes is awash in orchestra color with winds (flutes, piccolo, oboes, clarinets, bassoons), horns (trumpets, trombones, tuba), percussion (timpani, cymbals,snare), two harps and a wordless female choir. The instrumental separation just seemed more prominent with the Cybershaft in the mix.  This opinion remained consistant with just about all other orchestral pieces I played as well.
 
Any downsides? Well I will say that when I switched genres and listened to some down and dirty garage rock the Cybershaft seemed to have perhaps a little less bite than the Liveclock. The Cybershaft provided the extra bit of clarity/resolution but in some songs I'm not sure if that is the best thing.  A good example is when I first got a Koetsu cartridge and listened to the Stooges "Fun House". The Stooges IMO work best on a primal level when they are bashing you over the head with sonic sledgehammers. Instrumental nuance and line separation is not the game here. A thick and hard hitting wall of sound that kicks you in the gut is. What is interesting is that Hasegawa-san recommended a Rubidium clock if my primary musical interest was rock.
 
If you already own a Liveclock and listen to a lot of Classical/Jazz and acoustic music I would recommend the Cybershaft Premium OCXO as a nice addition. I think from a musical enjoyment perspective it provides value worth the asking price. 
 
How does it compare to other clock alternatives? Unfortunately, I can provide no guidance there as I have not heard others in my system. Therefore I can only stick with my direct personal experience which has been that (a) Antelope >Internal Mutec (and RedNet) and (b) Cybershaft/Antelope > Antelope by itself. At $1K I'm not presuming it is the bestest clock in the whole world. What does seem rather unique about the Cybershaft is that via it's direct sales model it seems to serve a pricing niche well below most others. As always, opinions made are mine alone via 50 year old ears and my setup/room. YMMV.

 

[jabbr]

Gldgte
Thank you for your interesting write up.

Inteiguing that a Rubidum 10MHz clock would have a different sound effect than an OCXO 10 MHz clock.
Did they give a further explanation for that effect?
How does their Limited-range of clocks fit in the musical spectrum of Rubidum vs OCXO?

Regarding the Antelope LiveClock, would you describe that as having a fast lock or slow lock when it comes to the speed of getting a lock to the incoming 10MHz signal?
Trying to get a sense of the design parameters of the Antelope clock.

Thx for the write up.

Cheers

 

[gldgate]

Gldgte
Thank you for your interesting write up.

Inteiguing that a Rubidum 10MHz clock would have a different sound effect than an OCXO 10 MHz clock.
Did they give a further explanation for that effect?
How does their Limited-range of clocks fit in the musical spectrum of Rubidum vs OCXO?

Regarding the Antelope LiveClock, would you describe that as having a fast lock or slow lock when it comes to the speed of getting a lock to the incoming 10MHz signal?
Trying to get a sense of the design parameters of the Antelope clock.

Thx for the write up.

Cheers

 
Yes, I was also a little surprised with Hasegawa-san's recommendation. I thought he would clearly steer me toward one clock or another. However, he said the clocks have a different "expression of sound" and provided me the recommendation based on type of music. What is interesting is that they do offer a box that contains both clock types. Someone over at the Audiogon forums (zephyr24069​) has this box and initially used the Rubidum only. When someone asked him about the OCXO he mentioned he had not tried it. When he did he came back and said he much preferred the OCXO to the Rubidium. Small sample size but thought I would pass it along. I thought about buying the combo box but ultimately decided against it because most of my listening is classical and the combo box has a much larger footprint (and space is becoming a premium for me).
 
The LiveClock seemed to lock onto the 10MHz signal quickly. Once connected the LED's switched instantaneously from "oven" to "Atomic". There is also an Atomic clock calibration mode within the Liveclock. Once you set up an Atomic Clock to the Liveclock and both devices are warmed up you go into Antelope SW and there is a calibration button. Hitting the button provides an initial number that represents the error of the LiveClock related to the Atomic clock. In my case this number was .2HZ. Hitting the calibrate number again compensates for this difference and the new number became 0Hz.
 
Hope that helps.

 

[somestranger26]

Here is one comparison of the OCXO vs RB version. This poster has the combo unit and tried both outputs.
 
Quote: https://forum.audiogon.com/discussions/rubidium-clock
 

Good Morning! In the way of any update,....the Cybershaft has well over a 1500 hours on it at this point (and the RB and OCXO clock outputs are equally broken in) and I've done a lot of testing over a wide variety of material (CD, XRCD, SACD) and alot of different recordings.  During this time I've spent several days at a time with the Cybershaft Rb and then the OCX outputs; over the past month I have found myself using only the OCX output circuit for 10Mhz 50ohm output as sonically, it is a much better match for the widest variety of recordings and things simply sound their best over and above what the Rb-based 10Mhz output circuit provides.  At this point, I think it is safe to say that assuming no other changes to the system, that I will be sticking with the current config and using the OCX output over the Rb for playback of 90-95% of all material.  Questions...please let me know.

 

[gldgate]

 
Yes, that is the thread and poster I was referring to. He emailed me today. He's still very happy with his purchase.
 
I also forgot to mention that even though Cybershaft's site is in Japanese, they did reply back to all my inquiries in English so I would not hesitate to contact them if anyone has any questions.  

 

[jelt2359]

Yes, that is the thread and poster I was referring to. He emailed me today. He's still very happy with his purchase.

I also forgot to mention that even though Cybershaft's site is in Japanese, they did reply back to all my inquiries in English so I would not hesitate to contact them if anyone has any questions.  

Thank you for your impressions. Any chance yet to try the 10 into the Mutec directly, and taking the liveclock out of the chain?

 

[Luckbad]

Can anyone clarify why you would ever need more than one word clock?

Meaning, why would you ever use a master clock to the Rednet then another clock/reclocker after it?

Aren't you discarding the original clock in something like the Mutec?

 

[somestranger26]

Can anyone clarify why you would ever need more than one word clock?

Meaning, why would you ever use a master clock to the Rednet then another clock/reclocker after it?

Aren't you discarding the original clock in something like the Mutec?

Mutec is not being used as a word clock here. It is only used as an SPDIF reclocker.

 

[gldgate]

Thank you for your impressions. Any chance yet to try the 10 into the Mutec directly, and taking the liveclock out of the chain?

 
I was curious and did do a quick compare (not extended time period) and much preferred sq with liveclock. However, given the newness of the cybershaft in my system I feel it best if I live with current setup for a while, get used to the new baseline and then try again.

 

[rb2013]

Looking at the phase noise numbers, this is a better buy than that Jackson Labs Fury that was posted a few pages back. Cybershaft Premium is rated -130dB/Hz at 10Hz and the Fury is -125dB/Hz. Both of these values blow away the Antelope 10MX, though, which is rated at only -87dB/Hz (100x as much noise).

The upcoming Mutec Ref 10 is rated for -140dB/Hz, and is rumored to cost 3000 Euro. If the Cybershaft sounds good, it's probably not worth spending so much on something like Ref 10. I think the diminishing return with such an upgrade would be massive.

Looking forward to your impressions.

Apples and oranges. The Ref10 and SRS PERF 10 are Rubidum clocks, the Cybershaft is OXCO. Just a 10Mhz OXCO. Relock the OXCO Antelope with another OXCO?

 

[rb2013]

Can anyone clarify why you would ever need more than one word clock?

Meaning, why would you ever use a master clock to the Rednet then another clock/reclocker after it?

Aren't you discarding the original clock in something like the Mutec?

These external 10Mhz clocks as a disciplining clock to internal clock, working to reduce the internal clocks variations. The phase noise numbers are not as important, as the internal clock is the final clocker.

This is completely different for an ext clock or spdif reclocker.

 

[jabbr]

These external 10Mhz clocks as a disciplining clock to internal clock, working to reduce the internal clocks variations. The phase noise numbers are not as important, as the internal clock is the final clocker.

....

I don't know if I would agree with this statement.
Phase noise is an indicator of oscilator short term stability and I would imagine this would be just as important for a 10MHz clock as it would be for a word clock.
The word clock is probably locked to the 10MHz signal by a PLL circuit as well. The word clock will then follow the fluctuations of a 10MHz clock signal in generating its word clock, how far it follows will be depending on the PLL circuit design parameters.

How do you view this, why would it for a 10 MHz signal not be important what its phase noise is? Why would it be decoupled?